chrishollandaise / spice-analysis

Statistical analysis on the relationship between spices and other variables using data from Dollar Street

Spice Analysis

Collage of family spices from Dollar Street. Photos by: Johan Eriksson, Lucy Forsyth, Isaiah Williams, Omar Hayana, Rosine Fidele, Joseph Ramirez, Raashi Saxena, and Alisa Sidorenko. Licensed under creative commons 4.0

---

Description

This is an accompanying repository for our paper that tried to answer the following statistical question: How Income & Location Impact the Number of Spices in a Household.

Technologies used for this analysis

• Python 3.8.3
• RStudio 1.3.1073

---

How to use

Before trying out our code, make sure to install some R dependencies: ggplot2 and RColorBrewer. For installation of our python dependencies, use terminal to cd into this folder, and use this command:

pip3 install -r requirements.txt

---

Sampling frame

Our sampling frame was created using code/appendix.py, which scraped family data from Dollar Street. It was last gathered on October 10, 2020. This frame has a total of 397 cases. Here is some information explaining the columns of this frame.

Variables	Type	Description
id	Identifier	A unique string of characters and integers.
name	Identifier	The family name.
income	Numeric	Income of family in dollars. Please check how these values were calculated by Dollar Street.
country	Categorical	The name of the country the family lives in.
continent	Categorical	The name of the continent the family lives in.
url	Identifier	The Dollar Street URL of the family.
slug	Identifier	A unique identifier found at the end of the family URL.
short_desc	Identifier	A short description about the family.

References

Dollar Street

By using Dollar Street family data, it helped answer our question. Please check them out!